
\begin{table}
\caption{Predicting nostalgic rhetoric (based on human codings) in 50 party manifestos. Dependent variable measures sentences coded as nostalgic by  both coders (Model 1), at least one of two coders (Model 2), the DistilBERT approach (Model 3), the Base Dictionary (Model 4), and GPT 3.5 (Model 5) for the same set of sentences.  Dependent variables are measured as the number of nostalgic sentences per 1,000 sentences, and z-transformed for comparability. Standard erorrs in parentheses.}
\begin{center}
\begin{scriptsize}
\begin{tabular}{l c c c c c}
\hline
 & M1 (Coding: =2) & M2 (Coding: >=1) & M3 (DistilBERT) & M4 (Dictionary) & M5 (GPT 3.5) \\
\hline
(Intercept)                     & $-0.10$     & $0.10$      & $0.08$      & $-0.05$      & $-0.05$     \\
                                & $(0.29)$    & $(0.26)$    & $(0.27)$    & $(0.24)$     & $(0.27)$    \\
Cultural Conservatism           & $0.37^{**}$ & $0.40^{**}$ & $0.38^{**}$ & $0.50^{***}$ & $0.40^{**}$ \\
                                & $(0.13)$    & $(0.12)$    & $(0.13)$    & $(0.11)$     & $(0.13)$    \\
Region: Northern E. (ref.: CCE) & $0.38$      & $0.05$      & $0.05$      & $-0.29$      & $0.13$      \\
                                & $(0.45)$    & $(0.41)$    & $(0.43)$    & $(0.38)$     & $(0.43)$    \\
Region: Southern E.             & $0.51$      & $0.52$      & $0.47$      & $0.79^{*}$   & $0.62$      \\
                                & $(0.39)$    & $(0.36)$    & $(0.37)$    & $(0.33)$     & $(0.37)$    \\
Region: Western E.              & $-0.05$     & $-0.49$     & $-0.39$     & $-0.03$      & $-0.15$     \\
                                & $(0.37)$    & $(0.34)$    & $(0.35)$    & $(0.31)$     & $(0.35)$    \\
\hline
R$^2$                           & $0.18$      & $0.30$      & $0.25$      & $0.41$       & $0.25$      \\
Adj. R$^2$                      & $0.11$      & $0.24$      & $0.18$      & $0.36$       & $0.18$      \\
Num. obs.                       & $50$        & $50$        & $50$        & $50$         & $50$        \\
\hline
\multicolumn{6}{l}{\tiny{$^{***}p<0.001$; $^{**}p<0.01$; $^{*}p<0.05$}}
\end{tabular}
\end{scriptsize}
\label{tab:reg_humancoded}
\end{center}
\end{table}
